Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Having some Issues with NDLR and VCV Rack
#1
Hey there, this is my first post here.

I got my NDLR a couple months ago but because of moving my office studio setup I didn't really get much time to experiment with the NDLR until the past week or so.

My setup consists of a bunch of various hardware, rack mounted synths and standalone keyboards, and I use VCV Rack in place of a traditional DAW. Hopefully there might be some people here with VCV experience. I like the modular workflow but am a little hesitant to actually jump into physical modular just because of cost (and addiction potential). I convert CV to Midi and send info out to all my gear using an iConnectivity midi hub, that has always worked without issue. I have used the NDLR a bit just hooked directly to my hardware and also no issues there.

My issues began to start when I wanted to route NDLR through VCV. Although I have some hardware I still use plenty of software sound sources, either VCV modules or VST instruments in VCV using the Host module. I'll start out just by saying that I know that some of these issues might be purely with how the VCV midi to CV modules work, but I'm hoping maybe some people have some solutions or tips.

The first issue I encountered was with Gate information, although I was able to get this somewhat workable. I did my research and I totally understand that NDLR doesn't actually send a "gate" just note on / note off midi messages. In VCV I've been able to somewhat successfully take either the "Gate" or "Retrig" outputs of the "Midi-CV" module in VCV rack and convert those pulses into functional gates that I can then use with my envelope generators. The problem is this tends to be very inconsistent and either I get missed notes or over time the "Gate" output on the Midi-CV module just appears to get stuck open. This happens both with Midi output into my iConnectivity hub, as well as Midi over USB. As far as I can tell even with other Midi-CV modules made by 3rd parties there isn't an output I can grab that is just looking at the Note On/Off signal, note on/off signals all seem to get parsed into v/oct + gate + velocity in VCV midi modules. For reference when testing this setup using Midi over USB I set each of the 4 parts to their own USB midi bus, each with their own separate channel. When using standard 5-pin midi I split 2 channels on A and 2 channels on B, and this leads into the next issue.

The 2nd issue is when using 5-pin din in this set up I get some really weird unwanted signals being sent all over the place. I have triple checked the routing both in the NDLR settings, and on my MIDI interface. Once I start playing more than 2 parts, it seems like garbage data starts getting randomly sent to every Midi channel. For example I can run Motif 1 and 2, but as soon as I start Drone random garbage pitch data starts getting output to Motif 1 and 2. This issue is circumvented by using USB for everything, which so far I can't see any downsides to. I originally switched over to using DIN because I was hoping that maybe I could get more usable gate signals that didn't drop notes, but I had more problems using DIN than just sticking to USB only. After a certain amount of time of getting this garbage message accumulation note data would just stop being sent to the Midi-CV modules no matter how many times I panic'd them out. The only way to fix it was to power off and on the NDLR, and in some cases I had to delete and re-add the Midi-CV modules in VCV rack (this leads me to believe this particular problem is a VCV issue, not the NDLR)

My final issue has to do with sending Midi CC values to the NDLR. In VCV I have a CV-CC module set up so that I can send CC's to CC 73 and 74 so that using VCV sequencers I can change the Key and Scale remotely to create an actual song structure. I also send pitch information (in the form of standard note on / off) to the NDLR to be able to sequence the Chord Degree / Type. The issue is that in VCV when I send the signal all these changes are not properly synchronized, so for example on the first beat of bar 16 of a song, I'd send values to change to E Minor Degree I. What happens a lot of the time is it will change the Key on the downbeat, but the Mode change happens a fraction of a second later (or sometimes not at all) which is very very noticeable in a musical sense. I know Midi (even over USB) has its limitations, but I feel like only sending 2 CC changes on their own dedicated control channel should be happening in a perceptibly simultaneous time. Instead I get a weird staggering effect that is easy to hear and sounds really bad.

I really want to use my NDLR with VCV to help me experiment and create more musical pieces without having to do it all with modules, but the Gate issues and the CC change issues make it almost unusable at times. Like I said I know some of this is pretty dependent on how VCV works (sending gate vs. note on/off), but I'm confused about the CC change issue. I also don't know if this is potentially an issue with just my unit too (I get a flickery screen sometimes and other graphical issues despite having not used the unit much). Also just to note I have had 0 problems I can think of using multiple other MIDI devices and controllers with VCV Rack, but none of them really are sending as much data as NDLR does, which might be contributing to the issue.

I've also included a screenshot of the NDLR template I set up in VCV Rack. From left to right the modules are CV to CC (Key/Mode) > CV to Midi (Degree Change) > 4x Midi to CC (Pad, Drone, Motif 1 & 2) > 3 modules set up to handle transposition from an external keyboard by NDLR (this works fine btw).
[Image: rTuxzm0.png]


If anyone has any experience using the NDLR with VCV Rack I'd really appreciate some input, or just to talk about some of these issues and what might be causing them.

Thanks

Just realized I made a small mistake with my issue regarding CC changes for Key and Mode. I can send the CC changes simultaneously and they work fine. It's when I try to also Send a Chord type / Degree at the same time as those with Note On/Off data that it takes an extra step before it changes instead of changing simultaneously. I have Pad Quantization off. and I don't think it has to do with any quantization settings in general because I'm not sending any other degree changes for a full 4 bars or more before sending the data to change the degree. It also is inconsistent, for example I have a 32 step pattern with 6 degree changes, sometimes all 5 will send in sync and just the last one is out of sync, sometimes its the 3rd one, sometimes it's none of them. Kinda frustrating to track down what is causing it. I also tried converting the sequencer gate into a trigger and sending the trig to the NDLR instead of the full 1 beat gate, but that didn't seem to make any difference.
Reply
#2
I uploaded a video highlighting the issue with sending degree changes simultaneously that might help clarify if I didn't explain it well.

Video example

You can see on the scopes that the CV change and Gate change are happening at the same time. However you can hear (and see later on the scope) that the pad is playing inconsistently and out of sync. I have Strum and Speed on the pad both set to 1 so it should just be playing a solid chord on the downbeat when the gate happens, but instead I get this staggered signal. I couldn't record the NDLR at the same time, but you can see the degree quickly change each time it happens.

So sending CC to 26 (for direct degree change) plays in sync with the key and mode changes, but when changing degree through CC there is no gate/trigger signal coming through the Pad or Drone outputs now. So unlike pressing the buttons on NDLR itself, the degree change happens but it doesn't trigger the pad chord essentially making the channel useless unless it is just left fully open. The drone "technically" works since it still changes notes on key changes, but it's not functioning like if you pushed the button directly (changing the note within the degree + key when pressed). Not sure if that is intentional or not.
Reply
#3
I can't tell whats happening from the video well enough to diagnose the issue with the pad part, but from what you describe with the Drone it might just not be set to one of the settings that follows the chord degrees. All of the Drone settings with 5 dots are meant to follow the chord degrees, while the ones before that play just the Root of the current Key, so that would make sense that they change with the Key, but not the chord degree changes.

To troubleshoot the Pad, can you try using that same setup being sent to other device? I'd like to see if it's an issue that persists outside of VCV rack. Does pressing the Chord Degree buttons cause the Pad to work properly but not CC control here? It doesn't sound like the timing is consistently off by the same amount of time, so this is probably not the solution, but check the settings pages and see if you can set pad quantize to off and see if that makes a difference.

Jesse
Reply
#4
(04-30-2021, 03:45 PM)Jesse Johannesen Wrote: I can't tell whats happening from the video well enough to diagnose the issue with the pad part, but from what you describe with the Drone it might just not be set to one of the settings that follows the chord degrees. All of the Drone settings with 5 dots are meant to follow the chord degrees, while the ones before that play just the Root of the current Key, so that would make sense that they change with the Key, but not the chord degree changes.

To troubleshoot the Pad, can you try using that same setup being sent to other device? I'd like to see if it's an issue that persists outside of VCV rack. Does pressing the Chord Degree buttons cause the Pad to work properly but not CC control here? It doesn't sound like the timing is consistently off by the same amount of time, so this is probably not the solution, but check the settings pages and see if you can set pad quantize to off and see if that makes a difference.

Jesse

Thanks for the reply Jesse,

I have been doing some more testing with my other hardware and everything on the NDLR works perfectly in terms of normal use. I'm aware of the Drone settings that either follow the root or the chord degrees, and I did have it set to the mode that is supposed to sync with degree changes. If I do everything through CC control (Key, Mode, Degree) there are no timing issues, but like I said the actual gate triggers don't come through for Pad or Drone. While hooked up to VCV and playing the same pattern with CC changes, pressing the buttons on the unit DOES properly send gates. So if for example I have strum turned on for the Pad channel, pressing the button for chord degree III properly sends the on/off info and it gets converted to gates/trigs in VCV, but a CC change to chord degree III doesn't. 

I don't really have any other hardware devices besides my PC to test the exact same scenario (Sending CC to control degree and see if it properly triggers Pad and Drone changes), but I could try in a different DAW besides VCV. But the Pad and Drone not retriggering on CC change does seem like it's something tied to the NDLR and not VCV, unless someone can chime in with experience controlling chord degree through CC using other sequencing hardware or DAW that doesn't have this behavior.

I do think that some of the issues I described in my first post have to do with the way VCV handles its Midi-CV conversion, because I have seen people talk about issues with gate/trigger when using the NDLR with their modular gear (stemming from the note on/off conversion as was mentioned previously). At this point though it's definitely more usable than when I started, really the only issue now is that CC changing the chord degree is not behaving the same way as just pressing the buttons on the unit itself. I'm not sure if pressing the button on the NDLR itself is sending more control information than just changing the CC does, but it seems that way.
Reply
#5
(05-02-2021, 12:38 PM)GreatSphynx Wrote:
(04-30-2021, 03:45 PM)Jesse Johannesen Wrote: I can't tell whats happening from the video well enough to diagnose the issue with the pad part, but from what you describe with the Drone it might just not be set to one of the settings that follows the chord degrees. All of the Drone settings with 5 dots are meant to follow the chord degrees, while the ones before that play just the Root of the current Key, so that would make sense that they change with the Key, but not the chord degree changes.

To troubleshoot the Pad, can you try using that same setup being sent to other device? I'd like to see if it's an issue that persists outside of VCV rack. Does pressing the Chord Degree buttons cause the Pad to work properly but not CC control here? It doesn't sound like the timing is consistently off by the same amount of time, so this is probably not the solution, but check the settings pages and see if you can set pad quantize to off and see if that makes a difference.

Jesse

Thanks for the reply Jesse,

I have been doing some more testing with my other hardware and everything on the NDLR works perfectly in terms of normal use. I'm aware of the Drone settings that either follow the root or the chord degrees, and I did have it set to the mode that is supposed to sync with degree changes. If I do everything through CC control (Key, Mode, Degree) there are no timing issues, but like I said the actual gate triggers don't come through for Pad or Drone. While hooked up to VCV and playing the same pattern with CC changes, pressing the buttons on the unit DOES properly send gates. So if for example I have strum turned on for the Pad channel, pressing the button for chord degree III properly sends the on/off info and it gets converted to gates/trigs in VCV, but a CC change to chord degree III doesn't. 

I don't really have any other hardware devices besides my PC to test the exact same scenario (Sending CC to control degree and see if it properly triggers Pad and Drone changes), but I could try in a different DAW besides VCV. But the Pad and Drone not retriggering on CC change does seem like it's something tied to the NDLR and not VCV, unless someone can chime in with experience controlling chord degree through CC using other sequencing hardware or DAW that doesn't have this behavior.

I do think that some of the issues I described in my first post have to do with the way VCV handles its Midi-CV conversion, because I have seen people talk about issues with gate/trigger when using the NDLR with their modular gear (stemming from the note on/off conversion as was mentioned previously). At this point though it's definitely more usable than when I started, really the only issue now is that CC changing the chord degree is not behaving the same way as just pressing the buttons on the unit itself. I'm not sure if pressing the button on the NDLR itself is sending more control information than just changing the CC does, but it seems that way.
I've been talking to the team and it sounds like the issue may lay with VCV's ability to process Note Off Messages sent Directly before a Note On Message. Since the NDLR doesn't keep track of gate length, each note is on until the next note is fired, at which point the NDLR sends first a Note off and immediately after, a Note On for the new note. This has shown not to play nice in some scenarios, with occasional hung notes (for me on my Ipad occasionally I experience this) and it seems that VCV rack was another example that has had some issues (I wasn't aware of that before). 
I can't guarantee that's what's happening here, but it sounds like a likely culprit. The good news is we're currently testing a fix for this and should hopefully have a solution to work with shortly. I'll keep you posted!
Jesse
Reply
#6
Thanks for the info Jesse, I appreciate it.
For the time being the device is still a lot of fun to use and get ideas with, but in the future it would definitely be great to get some more robust options or settings with regards to handling the note on/off information. I figured it was a combination of how NDLR sends its data and how VCV processes it, so good to know I wasn't that far off. Appreciate the responses, and if you need someone to help test beta firmware with VCV Rack let me know Wink
Reply
#7
“If anyone has any experience using the NDLR with VCV Rack I'd really appreciate some input, or just to talk about some of these issues and what might be causing them.”

It’s been a while since I’ve used NDLR with VCV and don’t recall having issues, albeit it was a while ago using an older FW revision. I skimmed through the post but didn’t see if you included that patch. However, as a thought, if you post the patch then I and/or others can confirm their experience as well.
Reply
#8
The audio interface for my VCV machine is being repaired, but I'm looking forward to getting it back and playing with the two together.
Reply
#9
(05-02-2021, 02:30 PM)GreatSphynx Wrote: Thanks for the info Jesse, I appreciate it.
For the time being the device is still a lot of fun to use and get ideas with, but in the future it would definitely be great to get some more robust options or settings with regards to handling the note on/off information. I figured it was a combination of how NDLR sends its data and how VCV processes it, so good to know I wasn't that far off. Appreciate the responses, and if you need someone to help test beta firmware with VCV Rack let me know Wink

I don't think there will be settings to choose from, any change will be global and permanent. The issue seems to be on the receiving end as Steve mentioned that there is no minimum time to wait between note on and off according to MIDI specifications, so gear should be able to process that without issue. We've just noticed the issue on more than one piece of gear, and figured we'd see if this solves it. 

I'll ask Steve regarding testing it and PM you if he's down. Otherwise it'll likely be posted soon.
Jesse
Reply


Forum Jump:


Users browsing this thread:
1 Guest(s)